Efficient Powertrain Control via Stochastic Signal Decomposition: Utilizing ARIMA Constituents for Real-Time knock detection in Internal combustion engines.

Knocking combustions in an Internal Combustion engine (ICE) are engine damaging combustions and hence reliable detection of every knocking event is very important. Engines typically rely on piezo-electric knock sensors to monitor structure-borne noise; which outputs a complex, continuous time series signal, thus differentiation of knocking vs knocking signal (signal to noise ratio) is difficult without computation intense signal processing such as Fast Fourier Transforms(FFT) or Wavelet transforms followed by manual calibration. In this research we propose an alternative to replace traditional knock detection with more reliable time-domain alternative using Autoregressive Integrated Moving Average (ARIMA) technique. Here we decompose the raw sensor signal into seasonality, trend and residue, and analyze the residue signal to differentiate knocking or non-knocking combustion, as the residual component is seen to retain abnormalities in the signal during knocking combustion. Further based on the amplitude of the residual, we can also easily classify the combustions as light, medium or heavy knocking events thus providing a precise and reliable detection.